GPU Comparison

The NVIDIA GeForce GT 740M and the NVIDIA GeForce 920M are both end-of-life mobile graphics solutions built on the same Kepler 2.0 architecture, yet benchmark data reveals a consistent performance advantage for the older GT 740M. Across the two available head-to-head tests, the GT 740M secures both wins, though the margin varies significantly depending on the workload.

Head-to-Head Benchmarks

The most decisive victory for the GT 740M comes in the Geekbench Vulkan test, where it scores 3459 against the 920M's 2849. This represents a 21.4% lead, a substantial gap that indicates the GT 740M handles modern graphics API workloads with considerably more efficiency. The Vulkan benchmark stresses draw calls and driver overhead, and the data shows the GT 740M's higher clock speeds translate into a clear advantage in this scenario.

In the Geekbench OpenCL test, the margin narrows considerably. The GT 740M scores 3974, while the 920M trails at 3725, yielding a 6.7% delta. OpenCL compute workloads often scale with raw shader throughput, and while the GT 740M still wins, the closer result suggests the 920M's architecture is not far behind in pure compute tasks. Still, a win is a win, and the GT 740M takes this test outright.

The average benchmark scores reinforce this trend. The GT 740M posts an average score of 3717, placing it in the 22nd percentile of all GPUs. The 920M, by contrast, averages 3287 and sits in the 20th percentile. The 430-point difference in average scores is consistent with the OpenCL result, showing that the GT 740M's advantage is not an artifact of a single test but a general performance characteristic.

Looking at the nearest rivals provides context for these scores. The GT 740M's average of 3717 is essentially tied with the NVIDIA Quadro 3000M, which scores 3718 with a 0% delta. It is also 0.6% ahead of the GeForce 825M (3694) and 1.9% ahead of the AMD Radeon HD 6770 (3649). The 920M's average of 3287 places it 0.9% behind the GeForce GT 730M (3316) and 1.4% behind Intel HD Graphics 530 (3332). This comparison shows that while both chips are entry-level performers, the GT 740M competes in a slightly higher performance tier.

The deltaPct values from the head-to-head data are the definitive measure. The GT 740M leads by 6.7% in OpenCL and 21.4% in Vulkan. No test shows the 920M ahead. The wins column is 2-0 in favor of the GT 740M, and there is no benchmark result in the entire FACT PACK where the 920M outperforms its older counterpart.

Architecture Differences

Both GPUs are built on the same fundamental architecture: Kepler 2.0, fabricated on a 28 nm process at TSMC. The transistor count is identical at 1,020 million, and the die size matches at 87 mm², resulting in the same transistor density of 11.7M per square millimeter. The underlying silicon is essentially the same, but the chips differ in designation, the GT 740M uses the GK208 chip, while the 920M uses the GK208B variant.

The core configurations are also identical. Both GPUs feature 384 shading units, 32 texture mapping units, and 8 raster output units. Neither chip includes ray tracing cores or tensor cores. Memory subsystems match perfectly: 2 GB of DDR3 on a 64-bit bus, delivering 14.40 GB/s of bandwidth. The memory clock is the same at 900 MHz with 1800 Mbps effective data rate.

The key architectural difference lies in clock speeds. The GT 740M has a base clock of 980 MHz and a boost clock of 1033 MHz. The 920M, in contrast, has a base clock of 954 MHz and a boost clock of 954 MHz, meaning it does not boost at all. This 26 MHz base clock disadvantage and 79 MHz boost clock disadvantage directly impact the 920M's compute rates. The GT 740M achieves a pixel rate of 8.264 GPixel/s and a texture rate of 33.06 GTexel/s, while the 920M manages 7.632 GPixel/s and 30.53 GTexel/s respectively. In FP32 compute, the GT 740M delivers 793.3 GFLOPS versus the 920M's 732.7 GFLOPS.

The physical form factor also differs. The GT 740M is listed as an MXM Module, a removable graphics card standard, while the 920M is an IGP (integrated graphics processor), typically soldered to the motherboard. Both have no power connectors and share a 33 W TDP, but the MXM form factor suggests the GT 740M was designed for upgradeable laptops, while the 920M was likely intended for more fixed configurations.

API support is identical: both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The bus interface is the same PCIe 3.0 x8. Display outputs are portable-device dependent for both. The GT 740M was released on June 19, 2013, as part of the GeForce 700M generation, while the 920M followed on March 12, 2015, in the GeForce 900M generation. The GT 740M's predecessor is the GeForce 600M and its successor is the GeForce 800M. The 920M's predecessor is the GeForce 800M and its successor is the GeForce 10 Mobile.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce GT 740M has an average benchmark score of 3717, while the NVIDIA GeForce 920M scores 3287. The GT 740M leads by 430 points.

Q: How large is the performance gap in Vulkan workloads?

A: In the Geekbench Vulkan test, the GT 740M scores 3459 against the 920M's 2849, resulting in a 21.4% advantage for the GT 740M.

Q: Are the core counts the same between the two GPUs?

A: Yes, both the GT 740M and the 920M feature 384 shading units, 32 TMUs, and 8 ROPs. The core configuration is identical.

Q: Do the GPUs share the same memory configuration?

A: Yes, both have 2 GB of DDR3 memory on a 64-bit bus with 14.40 GB/s bandwidth and a 900 MHz memory clock.

Q: What is the difference in boost clock speeds?

A: The GT 740M has a boost clock of 1033 MHz, while the 920M's boost clock is 954 MHz. The 920M does not boost above its base clock.

Q: Which GPU is positioned higher in the overall performance percentile?

A: The GT 740M is in the 22nd percentile of all GPUs, while the 920M is in the 20th percentile.

Specification Differences

| Specification | NVIDIA GeForce GT 740M | NVIDIA GeForce 920M |

|---|---|---|

| Chip | GK208 | GK208B |

| Generation | GeForce 700M | GeForce 900M |

| Base Clock | 980 MHz | 954 MHz |

| Boost Clock | 1033 MHz | 954 MHz |

| Pixel Rate | 8.264 GPixel/s | 7.632 GPixel/s |

| Texture Rate | 33.06 GTexel/s | 30.53 GTexel/s |

| FP32 Compute | 793.3 GFLOPS | 732.7 GFLOPS |

| Slot Width | MXM Module | IGP |

| Release Date | 2013-06-19 | 2015-03-12 |

| Predecessor | GeForce 600M | GeForce 800M |

| Successor | GeForce 800M | GeForce 10 Mobile |

| Average Benchmark Score | 3717 | 3287 |

| Percentile | 22 | 20 |

| Geekbench OpenCL | 3974 | 3725 |

| Geekbench Vulkan | 3459 | 2849 |

Where Each One Wins

The NVIDIA GeForce GT 740M wins in every benchmark category where data exists. In Geekbench OpenCL, it leads by 6.7%, demonstrating superior compute throughput for general-purpose workloads. In Geekbench Vulkan, its 21.4% lead is even more pronounced, making it the clear choice for applications that leverage modern graphics APIs. The GT 740M's higher boost clock of 1033 MHz versus the 920M's static 954 MHz directly contributes to its higher pixel rate (8.264 vs 7.632 GPixel/s), texture rate (33.06 vs 30.53 GTexel/s), and FP32 throughput (793.3 vs 732.7 GFLOPS). For any workload that depends on raw shading, texturing, or compute performance, the GT 740M is the stronger option.

The NVIDIA GeForce 920M does not win any benchmark test in the data. Its only advantages are qualitative: it is a newer product in the GeForce 900M generation, and its IGP form factor may be more suitable for thin-and-light laptop designs where a discrete MXM module is not practical. The 920M also has a later release date of March 12, 2015, which might imply better driver longevity, but the benchmark data shows no performance benefit from this. In terms of measurable performance, the 920M is consistently behind.

For users comparing these two for actual use, the GT 740M offers better compute and graphics performance across the board. The 920M's lower clocks mean it will produce lower frame rates in games and slower results in compute applications. The only scenario where the 920M might be preferable is if the motherboard integration of an IGP is required, since the MXM module of the GT 740M demands a compatible slot.

The Verdict

The data is unambiguous: the NVIDIA GeForce GT 740M outperforms the NVIDIA GeForce 920M in every measured benchmark. The GT 740M wins both head-to-head tests, holds a 430-point higher average score, and sits two percentile points higher in the global ranking. Its 21.4% Vulkan lead is particularly decisive, indicating that the GT 740M handles modern graphics APIs with far greater efficiency.

The 920M's only merits are its newer generation label and its integrated form factor. For a laptop manufacturer seeking a soldered GPU with a lower footprint, the 920M's IGP design might be simpler to implement. However, any user choosing between these two based on performance should select the GT 740M without hesitation. The 79 MHz boost clock advantage and the resulting 8.3% higher FP32 compute (793.3 GFLOPS vs 732.7 GFLOPS) translate directly into measurable performance gains.

The GT 740M also holds up well against its nearest rivals, matching the Quadro 3000M's average score and edging out the GeForce 825M by 0.6%. The 920M, meanwhile, trails the GeForce GT 730M by 0.9% and the Intel HD Graphics 530 by 1.4%, situating it at the very bottom of the entry-level tier. The verdict is straightforward: the GT 740M is the superior GPU by a meaningful margin, and the 920M's newer release date does not compensate for its lower performance metrics.